Skip to content
Hbench
hbench

Terminal-Bench 2.0 on Gemini 3.1 Pro Preview

89 tasks, 5 attempts, open network. Updated 25 Oct 2026.

Build a run for this board

Value

What a task cost each configuration.

$ per solved task

Lower is cheaper. Dollars per task divided by the pass rate.

Embed this chart

Not reported: tongagents, Ante, Gemini CLI

Tokens per task

Lower is leaner. Input plus output, cached input included.

Embed this chart

HarnessTokens per task
Gemini CLI860k
Not reported: tongagents, Ante

Agent time per task

Lower is faster. Agent start to finish; setup and tests left out.

Embed this chart

HarnessTime per task
Gemini CLI8.9 min×0.67
Ante12.5 min×0.94
tongagents0.8.0 · Reference13.2 min×1.00

Pass rate against time per task

Up and to the left is better. Not offered: tokens and dollars, which fewer than two rows report. Pass-rate axis starts at 50%.

Embed this chart

50%60%70%80%90%100%6.7 min8.3 min10.0 min11.7 min13.3 min15.0 minAgent time per taskMore time for a lower pass rate than a row on the frontiertongagents0.8.0 · 13.2 min · 80.2%Anteversion not recorded · 12.5 min · 73.7%Gemini CLIversion not recorded · 8.9 min · 59.8%
  • On the frontier: no other row has a higher pass rate for less time
  • Off the frontier
  • 95% interval of the pass rate
On the frontier, least time first: Gemini CLI, 8.9 min a task, 59.8%; Ante, 12.5 min a task, 73.7%; tongagents 0.8.0, 13.2 min a task, 80.2%.
  • 3 runs
  • Updated 25 Oct 2026
  • Method